Procurement Analytics

Procurement Analytics

Overview

Generate procurement analytical insights from structured data sources: spend analytics (category, vendor, timeline), KPI dashboards (order cycle time, approval turnaround, supplier performance), cost variance reports (budget vs. actual), supplier quality trending (inspection results, quality scores), and predictive analytics (risk prediction, price forecasting) based data, and supplier quality trending (inspection results, quality scores. Aggregates data from procurement systems (ERP, order management, supplier database) and generates visualizations and reports.

Announce at start: "I'm using the procurement-analytics skill to analyze procurement data and generate analytical insights."

When to Use This Skill

Trigger Conditions:

• Generating spend analysis reports (category breakdown, vendor concentration, trend charts)
• Creating procurement KPI dashboards (order cycle time, approval turnaround, supplier performance)
• Producing cost variance reports (budget vs. actual, variance causes)
• Tracking supplier quality trends (inspection results, quality scores over time)
• Building procurement performance reports for management review
• Generating export reports for audit or regulatory compliance

Prerequisites:

• Structured procurement data available (orders, approvals, invoices, delivery records)
• Time range defined for analysis
• Dimension of interest identified (category, vendor, project, time period)

Step-by-Step Procedure

Step 1: Analytics Type Selection

Select the type of analysis required:

Step 2: Data Source Identification

Identify data sources required for the analysis:

Step 3: Aggregation Pipeline

Execute data aggregation following the pipeline:

DATA EXTRACTION (source systems → structured dataset)
    ↓
FILTERING (time range, categories, vendors, projects)
    ↓
AGGREGATION (sums, averages, rates, trends)
    ↓
CALCULATION (variance, performance, ranking)
    ↓
VISUALIZATION DATA PREPARATION (chart-ready format)
    ↓
OUTPUT (report, dashboard, export)

Aggregation Examples:

Spend by Category:

SELECT category, SUM(amount) as total_spend, COUNT(*) as order_count
FROM orders
WHERE order_date BETWEEN start AND end
GROUP BY category
ORDER BY total_spend DESC

Supplier Delivery Reliability:

SELECT supplier_id, 
       AVG(CASE WHEN delivery_date <= expected_date THEN 1.0 ELSE 0.0 END) as on_time_rate,
       COUNT(*) as total_deliveries
FROM delivery_records
WHERE delivery_date BETWEEN start AND end
GROUP BY supplier_id

Step 4: KPI Calculation

Calculate procurement KPIs using standard definitions:

Step 5: Cost Variance Analysis

Calculate cost variance for orders and projects:

Cost Variance = Actual Cost - Budgeted Cost
Variance % = (Actual Cost - Budgeted Cost) / Budgeted Cost × 100

Variance Categories:

Step 6: Supplier Quality Trending

Track supplier quality performance over time:

Trend Analysis:

• Collect inspection results monthly (or per delivery)
• Calculate rolling average (last 12 months or last 20 deliveries)
• Identify trend direction (improving, stable, declining)
• Flag suppliers with declining quality trend

Quality Score:

Quality Score = (Passing Inspections / Total Inspections) × 100
Severity Weighted = Σ(inspection_result.weight) / Σ(inspection.weight) × 100

Step 7: Predictive Analytics (Future State)

Prepare for predictive analytics when ML models are available:

Step 8: Report Generation

Generate analytical report in requested format:

Success Criteria

• Analytics type correctly selected per requirements
• Data sources identified and accessed correctly
• Aggregation pipeline completes without data errors
• All KPIs calculated correctly with documented formulas
• Cost variance analysis identifies significant variances
• Supplier quality trending includes trend direction
• Report generated in requested format(s)
• Data accuracy matches source system values

Common Pitfalls

1. Incorrect Time Boundaries — Always use consistent date ranges across all data sources. Mixing fiscal and calendar periods causes comparison errors.
2. Double Counting — When aggregating orders and line items, avoid double counting (e.g., counting order total AND line items). Use consistent grain.
3. Ignoring Null Values — Null ≠ 0. Missing data should be flagged as null, not assumed to be zero. This affects averages and trends.
4. Stale Data — Always check last data refresh timestamp. Reports based on stale data mislead decision-makers.
5. No Baseline Comparison — KPIs without targets or baselines are not actionable. Always include the target or prior period comparison.

Cross-References

Related Skills

• procurement-data-extraction — Provides structured data for aggregation
• supplier-evaluation — Uses quality trending data for supplier scoring
• procurement-order-management — Provides order data for cycle time analysis
• procurement-compliance — Uses spend data for compliance monitoring

Related Agents

• Procurement Analytics Specialist (DomainForge) — Primary analytics owner, report generation
• Procurement Strategy Specialist (DomainForge) — Trend interpretation, recommendations
• Financial Compliance Specialist (DomainForge) — Cost variance analysis, budget compliance

Example Usage

Scenario: Generate Q1 2026 spend analysis and KPI dashboard for construction materials

1. Analytics Type: Spend analytics + KPI dashboard
2. Data Sources: Order records (Q1 2026), approval logs, delivery records, supplier database
3. Aggregation: Group orders by category (structural steel, concrete, electrical, mechanical)
4. KPIs: Order cycle time (4.2 days avg), approval turnaround (18 hours avg), on-time delivery (93%), budget compliance (102%)
5. Cost Variance: Structural steel +8% (above budget due to price increase), concrete -3% (under budget)
6. Quality Trend: Steel supplier quality declining (98% → 95% → 92% over 3 months)
7. Output: PDF report with executive summary, Excel export with detailed data

Performance Metrics

Target Performance:

• Data accuracy: >99.9% (matching source system aggregations)
• KPI calculation accuracy: 100% (correct formulas applied)
• Report generation time: <30 seconds (dashboard), <5 minutes (full report)
• Data freshness: <1 hour for dashboard, <24 hours for report